Aggregate escape counts across all f-string segments in normalize_fstring_quotes

[bysiber]

In normalize_fstring_quotes, the loop that checks whether quote normalization would introduce more escaping overwrites orig_escape_count and new_escape_count on each iteration:

for middle, new_segment in zip(middles, new_segments, strict=True):
    orig_escape_count = middle.value.count("\\")
    new_escape_count = new_segment.count("\\")

if new_escape_count > orig_escape_count:
    return middles, quote  # Do not introduce more escaping

After the loop finishes, both variables only hold the counts from the last segment. For an f-string with multiple text segments (e.g. f'text1{expr}text2'), a newly-introduced escape in an earlier segment is ignored entirely.

This changes the loop to aggregate the counts across all segments:

orig_escape_count = 0
new_escape_count = 0
for middle, new_segment in zip(middles, new_segments, strict=True):
    orig_escape_count += middle.value.count("\\")
    new_escape_count += new_segment.count("\\")

Note: this function is currently only referenced in a commented-out block in linegen.py, so there's no runtime impact today. But the fix is worth having before f-string quote normalization is enabled.

[github-actions]

diff-shades results comparing this PR (9846ad2) to main (5518079):

--preview style: no changes

--stable style: no changes

---

What is this? | Workflow run | diff-shades documentation